Translation memory can be considered as a feature which is available in computer-based translation systems. It is providing an excellent assistance to the translation process. In fact, the translation memory provides an ability for the translator to go ahead and re-use any of the textual segments, which were translated before.

In simple terms, the translation memory is a database which records all the information that the translator is working on. Depending on the nature of translation work, this can be small or large. The information stored in the database are based upon target sentence or source sentence concepts. For example, if a new sentence is coming in and the database has a similar entry, it would then show that sentence to the translator. The translator will be able to pick this and proceed with the translation work.

As you can see, translation memory is providing an excellent assistance to a translator. That’s because a translator doesn’t need to look for the reference materials. They will be recommended to the translator while he is proceeding with the job. This will help the translator to save both money and time when getting the translation work done. It can also be used and recalled partially or completely at a fraction of the time and cost, instead of translating the same content again from the scratch.

How does translation memory work?

Now you have a basic understanding of what translation memory is. With that in mind, it is worth to take a look at how translation memory works. Software tools are available to store translation memory when a translator is working on content. If a new piece of content that the translator translates matches perfectly well with a piece of content that is already stored in the database, the text will be extracted instantly. Then it will be placed in the target file.

When a translator is working on the source file, the software would look for matches in the background. If a match is detected, it will be suggested. The translator is then provided with the ability to make a decision and determine whether he is going to accept the suggestion or not. If the translator believes that the translated piece of content is not accurate or it doesn’t match, he is provided with the freedom to translate it from the scratch. Hence, there is a possibility to end up with a perfect outcome with the assistance of translation memory.

When should you use translation memory?

Any person who is localizing content will be able to use translation memory. It helps to translate content from one language to another with minimum hassle. When you are translating a document which has a high level of repetition, translation memory will be able to provide an excellent assistance to you with getting the job done. This will help you to complete the translation within a short period of time as well.

You should also keep in mind that the translation memories are extremely helpful when translating content from a text. That’s because a large number of companies are currently working on Content Management Systems in order to manage the information. With the assistance of a CMS, it is possible for blocks of texts to be edited or translated, instead of working on an entire document. In addition, the documents can be published in a variety of formats as well. With the assistance of translation memory, this job can be done within a short period of time. In addition, it is possible to end up with more consistent results as well.

Even if a translation memory is not used, the dedicated translation environment has the ability to provide assistance to you with extracting text from a source file. Then you will be able to edit the source file and then focus on localizing the text. As a result, you don’t need to worry too much about the tags. For example, if you are working on a HTML document, you will have to face some issues with the tags. However, editing the source file can keep you away from that hassle.

Is translation memory the same as machine translations?

Most of the people confuse in between the two terms translation memory and machine translations. The machine translations are completely automated. They are faster than translation memory. That’s because a person still has to work with translation memory to get a document translated.

However, the chances of ending up with mistakes is high when you use machine translation. In other words, the accuracy of the translation document will be somewhere in between 50% to 70%. Therefore, it is not similar to translation memory and it cannot be used to replace translation memory.

This article first appeared on Agato Translation Blog we find it worth sharing with you: